Cor Compare
2022-08-26
Corilation Compare!
I have run four separate experiments and now I will compare their different correlations to see if I can tell the difference between each of the four experiments.
Experiments:
- [A] tall map with 5 sections
- [B] tall map with 7 sections
- [C] tall map with 5 sections no heritability
- [D] tall map with 5 sections no interaction
I am investigating the correlation calculations. I predicted that though co-evolution (interaction & evolution) species’ phenotypes would become spatially correlated. I saw some very high correlations (around 0.5) in some of these simulations. To see if there is a difference between the correlations in my experiment I will create some comparative histograms.
Predictions:
- 5 sections & 7 sections will have similar correlation distributions
- no heritability & no interaction correlation distributions will look very different from 5 sections & 7 sections
Comparing the Mean
## # A tibble: 4 × 2
## experiment name
## <chr> <dbl>
## 1 A 0.0451
## 2 B 0.0441
## 3 C -0.0000572
## 4 D 0.000542## [1] "One mean per GA"## [1] "Anovas from mean of A and B"## Df Sum Sq Mean Sq F value Pr(>F)
## exp 1 0.00003 0.0000303 0.014 0.906
## Residuals 30 0.06392 0.0021308## [1] "Anovas from mean of A and C"## Df Sum Sq Mean Sq F value Pr(>F)
## exp 1 0.01710 0.017097 15.3 0.000486 ***
## Residuals 30 0.03352 0.001117
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## [1] "Anovas from mean of A and D"## Df Sum Sq Mean Sq F value Pr(>F)
## exp 1 0.01666 0.016657 4.433 0.0437 *
## Residuals 30 0.11271 0.003757
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## [1] "Anovas from mean of C and D"## Df Sum Sq Mean Sq F value Pr(>F)
## exp 1 0.00000 0.0000029 0.001 0.974
## Residuals 30 0.07922 0.0026408## [1] "One mean per rep"## [1] "Anovas from mean of A and B"## Df Sum Sq Mean Sq F value Pr(>F)
## exp 1 0.000 0.000035 0.004 0.948
## Residuals 124 1.012 0.008160## [1] "Anovas from mean of A and C"## Df Sum Sq Mean Sq F value Pr(>F)
## exp 1 0.0649 0.06486 14.48 0.000221 ***
## Residuals 125 0.5600 0.00448
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## [1] "Anovas from mean of A and D"## Df Sum Sq Mean Sq F value Pr(>F)
## exp 1 0.0631 0.06315 5.574 0.0198 *
## Residuals 125 1.4161 0.01133
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## [1] "Anovas from mean of C and D"## Df Sum Sq Mean Sq F value Pr(>F)
## exp 1 0.0000 0.000012 0.002 0.967
## Residuals 126 0.8566 0.006798I want to compare the different cor values from the simulations, but i also want to look at it simulation by simulation.
I think trials within a simulation experiment (A, B, C, D) should be the same (same mu and mu effect size).
They should also be different between different GA combinations
same GA should be different in the different experiments
##
## Wilcoxon rank sum test with continuity correction
##
## data: GA1_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno and GA1_cor_tall_7_file$mean_newt_pheno_By_mean_snake_pheno
## W = 5.0337e+10, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0##
## Wilcoxon rank sum test with continuity correction
##
## data: GA1_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno and her_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## W = 5.9204e+10, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0##
## Wilcoxon rank sum test with continuity correction
##
## data: GA1_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno and inter_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## W = 2.886e+10, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0##
## Wilcoxon rank sum test with continuity correction
##
## data: GA1_cor_tall_7_file$mean_newt_pheno_By_mean_snake_pheno and her_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## W = 5.9487e+10, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0##
## Wilcoxon rank sum test with continuity correction
##
## data: GA1_cor_tall_7_file$mean_newt_pheno_By_mean_snake_pheno and inter_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## W = 2.8987e+10, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0##
## Wilcoxon rank sum test with continuity correction
##
## data: her_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno and inter_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## W = 2.6368e+10, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0##
## Welch Two Sample t-test
##
## data: GA1_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno and her_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## t = 104.82, df = 476795, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 0.04447615 0.04617107
## sample estimates:
## mean of x mean of y
## 0.0451916473 -0.0001319622##
## Welch Two Sample t-test
##
## data: GA1_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno and inter_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## t = 63.878, df = 298681, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 0.04327940 0.04601935
## sample estimates:
## mean of x mean of y
## 0.045191647 0.000542273##
## Welch Two Sample t-test
##
## data: GA1_cor_tall_7_file$mean_newt_pheno_By_mean_snake_pheno and her_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## t = 117.61, df = 533338, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 0.04347920 0.04495287
## sample estimates:
## mean of x mean of y
## 0.0440840761 -0.0001319622##
## Welch Two Sample t-test
##
## data: GA1_cor_tall_7_file$mean_newt_pheno_By_mean_snake_pheno and inter_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## t = 65.423, df = 257034, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 0.04223737 0.04484624
## sample estimates:
## mean of x mean of y
## 0.044084076 0.000542273##
## Welch Two Sample t-test
##
## data: her_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno and inter_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno
## t = -1.0904, df = 198623, p-value = 0.2755
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.001886185 0.000537715
## sample estimates:
## mean of x mean of y
## -0.0001319622 0.0005422730##
## Welch Two Sample t-test
##
## data: GA1_cor_tall_file$mean_newt_pheno_By_mean_snake_pheno and GA1_cor_tall_7_file$mean_newt_pheno_By_mean_snake_pheno
## t = 2.2261, df = 609282, p-value = 0.02601
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 0.0001324152 0.0020827274
## sample estimates:
## mean of x mean of y
## 0.04519165 0.04408408
next step is to count the high and the lows
like make a lower limit like must be above 0.3
## # A tibble: 4 × 2
## `get(whichgroup)` n
## <chr> <int>
## 1 A 16969
## 2 B 10650
## 3 C 690
## 4 D 16138## # A tibble: 4 × 2
## `get(whichgroup)` n
## <chr> <int>
## 1 A 25862
## 2 B 15321
## 3 C 1352
## 4 D 31031## # A tibble: 4 × 2
## `get(whichgroup)` n
## <chr> <int>
## 1 A 4616
## 2 B 2249
## 3 C 16
## 4 D 7453## # A tibble: 4 × 2
## `get(whichgroup)` n
## <chr> <int>
## 1 A 456
## 2 B 230
## 3 C 0
## 4 D 417
